Wearable closed-loop TUS system and method thereof

Abstract

The invention discloses a wearable closed-loop TUS system and a method thereof. A TUS (Transcranial Ultrasound) system with a matrix array transducer in a wearable form is disclosed. The TUS system uses ASIC (Application Specific Integrated Circuit) and MEMS (Micro Electro Mechanical System) technologies to achieve power, size and weight reduction, allowing the device to be worn. In embodiments, the TUS system can reliably find the target anatomical structure and remain locked to the target throughout use. For closed-loop operation, all real-time tasks are locally controlled within the TUS system. Protective measures enable a subject to use the system at home and clinics for medical treatment and health use. A method for using the system is disclosed.

Description

Technical Field

[0001] The present invention relates to a transcranial ultrasound system (TUS) as a wearable health device that can operate in clinical and non-clinical environments. Background Art

[0002] TUS systems help treat several types of psychiatric disorders, but these systems do not guarantee that ultrasound stimulation reaches the anatomical target accurately. The curvature of the skull and variations in skull thickness obscure ultrasound (US) delivery to the desired location, especially for small and deep targets such as the amygdala.

[0003] Matrix array systems can solve the problem of poor alignment by acquiring volumetric data sets of brain anatomy. Current solutions are targeted at clinically controlled uses of treatment of diseases defined by clinicians. It is necessary to expand the capabilities of TUS systems as health devices for individuals who want to improve their health despite no medical diagnosis. In addition, home medical treatment controlled by physicians has cost and convenience advantages over clinical procedures. For this reason, TUS systems need to be in wearable form. In addition to ensuring that the wearable TUS system is safely and correctly aligned with the correct anatomical target, a large reduction in cost, power, size and weight is also required to make the system in a wearable form factor. Summary of the invention

[0004] A transcranial ultrasound system, namely a TUS system, comprises: a headband device, which comprises one or more transcranial ultrasound sheets, namely TUS sheets, wherein each of the TUS sheets further comprises a plurality of transducer elements connected to an integrated circuit; and a computing device, which wirelessly communicates with the headband device, wherein the computing device generates a stimulation signal specification and sends the stimulation signal specification to the headband device, wherein the TUS sheet outputs a stimulation signal via the headband device.

[0005] A method for a transcranial ultrasound system (TUS) system, comprising: placing a headband device on the scalp of a subject, the headband device comprising a plurality of transcranial ultrasound sheets (TUS sheets); obtaining TUS parameters from a memory, the TUS parameters being used to specify an operation mode of the TUS sheets; obtaining skull characterization data from the TUS sheets based on the TUS parameters; calculating skull aberration correction parameters based on the skull characterization data; adjusting the TUS parameters based on the skull aberration correction parameters; and generating a stimulation signal based on the adjusted TUS parameters, wherein the TUS sheets transmit the stimulation signal via the headband device. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 A wearable TUS system 100 is shown according to an embodiment.

[0007] Figure 2 is an exemplary block diagram of a TUS wearable device 120 used in the system 100 .

[0008] Figure 3 An example of a tile 240 including a PMUT MEMS array 250 and an ASIC 260 is shown.

[0009] Figure 4 An example ASIC 260 layout is shown organized in a 16×8 grid.

[0010] Figure 5 An example layout of one cell of a MEMS 250 die is shown.

[0011] Figure 6 An exemplary block diagram of ASIC 260 is shown in accordance with an embodiment.

[0012] Figure 7 An exemplary block diagram of an analog block 610 in ASIC 260 is shown.

[0013] Figure 8 An exemplary functional block diagram of digital block 620 is shown.

[0014] Fig. 9 An exemplary block diagram of a DWE 810 is shown.

[0015] Fig.10 is an exemplary method 1000 for closed-loop operation of a TUS wearable device 120 . DETAILED DESCRIPTION

[0016] A TUS (transcranial ultrasound) system in a wearable form factor with a matrix array transducer is disclosed. The TUS system uses ASIC (application specific integrated circuit) and MEMS (micro-electromechanical systems) technology to achieve reductions in power, size, and weight, allowing the device to be worn. In an embodiment, the TUS system can reliably locate target anatomical structures and remain locked to the target throughout use. For closed-loop operation, all real-time tasks are controlled locally within the TUS system.

[0017] In the present invention, there are two closed loops. One loop is that the guided measurements continuously update the wearable device's knowledge of the location of one or more stimulation targets. This knowledge is used to modify the direction and focus of the stimulation beam so that the target is locked even when the subject moves. The second loop is that the timing of the stimulation can be controlled by information derived from the EEG measurements. Safeguards enable the subject to use the system at home for medical treatment and health purposes. Methods for using the system are disclosed.

[0018] The present application discloses a TUS system in a wearable form with a matrix array transducer. A two-dimensional rectangular array of transducers is also known as a matrix array transducer. By utilizing the present invention, power, size and weight can be drastically reduced, making wearable devices possible. In implementation, ASIC (application-specific integrated circuit) and MEMS (micro-electromechanical system) technology achieve this reduction, allowing the device to be worn. In implementation, a MEMS die sits on top of the ASIC in a vertically interconnected manner. In an embodiment, the MEMS device is inverted so that the active surface of the MEMS device faces the active surface of the ASIC. The acoustic output comes from the back side of the MEMS die at the top of the stack. A cavity is etched into the back side of the MEMS die using a process such as deep reactive ion etching (DRIE).

[0019] In an embodiment, the TUS system can reliably find the target and remain locked to the target throughout use. The TUS system sends and receives sound (ultrasound) in various directions, thereby creating a "beam" in sequence. The beam can be controlled by a parameter set, which is stored in a local memory (e.g., in a table format) and is referred to as a "beam list" in this application. When a "start" signal is received (via an application (app) operating on a subject device such as a smart phone), all real-time tasks can be locally controlled (i.e., closed-loop operation) within the wearable system. The protection measures in the system ensure that the TUS system can be used at home or in a clinic under a treatment plan defined by a clinician. In different embodiments, the TUS system is used by subjects who do not have any medical diagnosis but use the TUS system to improve their health. In a health mode or application, the subject can use various preset schemes. In an embodiment, the TUS system includes physiological measurement capabilities using EEG (electroencephalogram), EOG (electrooculogram), EMG (electromyogram) or ECG (electrocardiogram) subsystems. In an embodiment, the system includes algorithms or machine learning functions for compensating for acoustic effects of the skull and EEG biomarker detection. In an embodiment, the TUS system can be connected to a cloud server where the machine learning functions can be updated based on cloud computing.

[0020] The term "drum" refers to a single MEMS structure having a vibrating suspended membrane actuated by a piezoelectric film.

[0021] The term "element" refers to an autonomous acoustic structure consisting of several "drums". All drums in an element are wired in parallel within the MEMS die. The ASIC electronics are connected to the entire element, not to individual drums.

[0022] The term "cell" refers to a group of "elements" whose electronics are repeated across the ASIC die. In one implementation, a cell controls the operation of 5 elements including 1 transmitter and 4 receivers.

[0023] In an embodiment, the system can operate in three modes: channel, beam, and stim. Channel mode operation can be used for skull characterization. Human skulls differ in thickness, sound velocity, and curvature, thereby distorting the ultrasound delivery to the desired target through attenuation, refraction, diffraction, and reflection. Channel mode is used to characterize the effect of the skull on the ultrasound transmission of all ultrasound transducer elements. The result of this operation is used to compensate for the effect of the skull on all ultrasound transducer elements. Beam mode operation can be used to draw a map of brain structure and anatomical characterization and to align brain structure and anatomical characterization. In an embodiment, the corrected element delay and amplitude obtained in the channel mode ensure that the element is correctly aligned with the desired anatomical structure. Stimulation (Stim or stimulation) mode is used for treatment. Table 1 below summarizes the operation of the three modes.

[0024] Table 1

[0025]

[0026]

[0027] There are differences between transmit beamforming operation and receive beamforming operation in beam mode. These differences mean that different parameters in the beam list control the transmit beamforming operation and the receive beamforming operation, as detailed below:

[0028] In the receive beamformer, time delays are continuously applied to the channel data to bring the image into focus at each point (dynamic receive focus). The schedule of delay variations is parameterized in the beam list as the depth of each channel increases.

[0029] • Adding together the data received from all transducer elements after applying the appropriate delays. This coherent summing can be done across slices by "chaining" the slices together.

[0030] • In the transmit beamformer, in the simplest mode of operation, there is a single set of delays, one for each channel, which creates a focus at a fixed depth.

[0031] • More advanced transmit beamforming for guidance may involve plane wave transmissions at different angles, from which data may be retrospectively combined to produce a dynamic transmit focus. In such a confocal system, both the transmitter and receiver are focused at various points in the image.

[0032] Figure 1 A wearable TUS system 100 according to an embodiment is shown. Figure 1 The subject 110 wears the TUS headband 120 (or TUS device) in a suitable position to ensure proper contact with the scalp of the subject 110. Other wearable sizes may be a helmet covering most of the scalp, a half helmet, a half headband, a flexible cap, and a mask.

[0033] The TUS system 100 includes a secure network device 130. All communications with the TUS system pass through the secure network device 130. This allows for authentication and encryption of all communications, tamper detection, protection against side-channel attacks, and isolation of the subject device 140 and the Internet from the operation of the TUS wearable device 120.

[0034] In a preferred embodiment, the headband 120 communicates with the secure network device 130 using a wireless network 170 such as Wi-Fi or Bluetooth. In an embodiment, data from the ASIC 120 is encrypted before it reaches the ASIC pins, making unauthenticated access to the subject's measurements very difficult. The FPGA 210 or microcontroller 290 can perform other encryption and decryption. In different embodiments, a wired network (Ethernet, USB, etc.) can be used for communication. In order to maintain high security, the subject 110 cannot install any software on the secure network device 130. The secure network device 130 will only allow secure commands to enter the system 100 to ensure the safe use of the wearable device 120 and the safety of the subject 110. The network device 130 communicates with the subject device 140 such as a smart phone, tablet computer, laptop computer, smart watch, computer, etc. using a wireless protocol such as Wi-Fi or Bluetooth. The subject 110 can use the application on the device 140 to control the operation of the system 100. Messages or other communications (data) from the system 100 are displayed via an application on the device 140. Both the network device 130 and the subject device 140 are connected to the cloud server 150 via the Internet 180, allowing authentication of the network device 130 and the subject device 140. The cloud server 150 enables non-real-time aspects of the system 100 such as updates to the machine learning aspects of the system, management of subject data, and over-the-air software updates. Clinicians using the portal 160 on the server 150 can monitor treatment and change their course of treatment. In addition to the ultrasound transducer, the headband 120 can also include an EEG subsystem. The EEG subsystem allows monitoring of the physiological efficacy of the system (such as the effect on the brain, etc.). The subject device 140 receives EEG and RX data. In different implementations, the system 100 includes at least one system with physiological measurement capabilities such as EOG (electrooculogram), EMG (electromyogram), or ECG (electrocardiogram).

[0035] In an implementation, the secure network device 130 includes a microprocessor that supports the necessary wireless protocols (Wi-Fi, Bluetooth, etc.). It translates high-level instructions from the subject device 140 or the cloud server 150 into low-level commands to the TUS device or headband 120. Instructions may include commands such as "start", "stop", etc., target location, treatment details, stimulation parameters, etc. Other non-real-time tasks are managed by the device 130. These tasks include authentication, encryption, tamper detection, protection against side-channel attacks, etc. ASICs, off-the-shelf processors, or FPGAs can be used to implement the device 130. In different implementations, the functionality of the device 130 is incorporated into the wearable device 120.

[0036] Figure 2is an exemplary block diagram 200 of a TUS wearable device 120 used in the system 100. Figure 2 , the device 120 consists of a real-time (RT) control device 210. In a preferred embodiment, the control device 210 is implemented using an FPGA. Other implementations of the control device 210 using an ASIC, a microprocessor, etc. are possible. The RT control device 210 communicates with the security network device 130 using a wireless protocol such as Wi-Fi or Bluetooth or a wired interface such as USB.

[0037] A plurality of sheets 240 are attached to the RT control device 210. Each sheet includes a MEMS device 250 (or a MEMS die) and an ASIC 260 (an ASIC die or an IC die). The MEMS device 250 is composed of a two-dimensional pMUT (piezoelectric micromachined ultrasonic transducer) array transducer and is controlled by an ASIC 260. Other technologies such as cMUT (capacitive micromachined ultrasonic transducer) or bulk piezoelectric transducer can also be used. In the block diagram 200, only three sheets to be connected are shown. However, the device 120 can be composed of many such sheets positioned on different head parts. In the case where the interconnection between the sheets is achieved by a flexible circuit, the sheets can be held in place by a flexible cap. Sheets, FPGAs and other electronic components and physiological sensors can be mounted on a flexible member arranged in the cap.

[0038] Figure 3 An example cross section of a sheet 240 including a pMUT MEMS array 250 and an ASIC 260 is shown. Figure 3 , one or more vertical connectors 320 are used to connect the MEMS 250 and the ASIC 260 . Figure 3 A MEMS silicon die 250 is illustrated showing the backside etch at the top side and a vibrating suspended membrane (referred to in this application as a drum) above a piezoelectric layer 310. The piezoelectric layer 310 is a sandwich including a layer of piezoelectric material such as PZT (lead zirconate titanate), AlN (aluminum nitride), or AlScN (aluminum scandium nitride), etc., with metal electrodes above and below it. Figure 3 Also shown is a propagation or coupling medium 340 (eg, gel) for ensuring proper acoustic coupling from the MEMS drum to the scalp of the subject 110. Figure 3 In FIG. 2 , the scalp of the subject 100 is positioned above the sheet 240 , and thus the tagged ultrasonic waves are guided into the head.

[0039] Return to reference Figure 2, a memory 220 is attached to the RT control device 210. The memory 220 contains a list of beam parameters. In an embodiment, the beam parameters are stored in a table format. The memory 220 may be an external DRAM, SRAM, EEPROM, memory within an FPGA, etc. The RT control device 210 reads the beam parameters from the memory 220 and uses the beam parameters to control the operation of the slice 240. The beam parameters may be downloaded to the memory from a storage device outside the device during the boot process.

[0040] Tiling or RX Linking

[0041] Multiple slices 240 may be connected. Each ASIC 260 within a slice 240 may accept a digital input stream 280 from an adjacent ASIC. After appropriate delay and amplitude adjustment, the digital input stream 280 is summed with the beamformed RX data generated by the MEMS device 250 connected to the ASIC. The summed result is passed to the ASIC of the next slice. The output 285 of the last slice 240 is sent to the RT control device 210.

[0042] EEG Subsystem

[0043] In an embodiment, an EEG subsystem 230 is included in the system 100. In an embodiment, the EEG subsystem 230 can communicate directly with the safety network device 130. The EEG subsystem 230 is used to monitor the effect of the treatment by monitoring the EEG signal for biomarkers. Care is taken to ensure that there is no mutual interference between the EEG subsystem 230 and the ultrasound operation. This is feasible because EEG data is very slow compared to ultrasound data: the frequency range of EEG data is about 1-100Hz, while ultrasound data is in the low MHz range. Typically, the number of EEG channels will be in the range of 3 to 64. The EEG operation involves passive reception of scalp electrical signals. This means that the scalp electrical signals can be directly received by the microcontroller 290 (labeled as μC) and do not need to be connected to the RT control device 210. The safety network device 130 also includes a wireless controller 295. EEG requires signal conditioning circuitry (not shown) such as analog amplification and signal filtering, which can be implemented using commercially available devices. Alternatively, and because the number of packages is advantageously reduced, these electronic devices can be implemented in the mixed signal ASIC 260.

[0044] Figure 4 An example ASIC 260 layout 400 is shown in a 16×8 grid. Figure 4, the layout 400 consists of 128 cells arranged in a 16×8 grid. In an embodiment, the cell consists of 1 transmit element and 4 receive elements. The analog functions of the cell include 4 LNAs (low noise amplifiers), 8 mixers to produce in-phase and quadrature band outputs for each receive element, 2 analog summers (summers), and 2 A / Ds (analog-to-digital converters). For the transmitter, a level shifter, output stage, and digital logic are required. Preferably, the functional digital pads on the ASIC 260 are only located on two sides, which helps to reduce the gap between the transducers. This improves the transmit and receive performance.

[0045] MEMS

[0046] Figure 5 An example layout of a unit cell of a MEMS 250 die is shown. Figure 5 In an embodiment, the MEMS 250 die is composed of 4 receiving sections 510 and a centrally placed transmitting section 520. As shown, the transmitting section 520 and the receiving section 510 are separated, which simplifies the implementation in the ASIC 260 by removing the need for any transmit / receive switching logic. Each of the 5 sections is connected to the ASIC 260. The receiving section 510 is composed of several "drums". The transmitting section 520 is composed of several "drums". In implementation, the number of drums for each RX and TX can be fine-tuned. Multiple drums are connected in parallel on the MEMS die to form the transmitting and receiving elements. In an embodiment, the receiving section 510 can operate at a different frequency than the transmitting section 520 (for example, the RX frequency is twice the TX frequency (2×TX frequency)). The example layout shown allows the receiving elements to be packaged more finely to allow the receiving section 510 to operate at a higher frequency than the transmitting section 520. Other layouts with different numbers of receiving sections 510 and transmitting sections 520 and in different geometric arrangements are possible.

[0047] ASIC

[0048] Figure 6 An exemplary block diagram of an ASIC 260 cell according to an embodiment is shown. Figure 6 , the ASIC 260 cell consists of an analog block 610 and a digital block 620. The area of ​​the ASIC cell can match the area of ​​the MEMS cell. If the ASIC cell area does not match the MEMS cell, an interposer can be placed between the two dies to enable connections from the component to the ASIC to receive inputs and send outputs.

[0049] Simulation Block

[0050] The analog block 610 is connected to the MEMS 250 cell using the interconnect 320. The analog block 610 converts the analog signal received from the MEMS 250 into a digital signal and provides it to the digital block 620. The transmit section inside the analog block 610 takes an input signal from the digital block 620 that has the correct timing and pulse train for one of the three types of operations in Table 1. The analog block 610 converts the input signal into an analog signal to drive the MEMS 250.

[0051] Figure 7 An exemplary block diagram of an analog block 610 in a cell of ASIC 260 is shown. Figure 7 , for the receiver, the receive analog function for RX0 MEMS 510 includes LNA 720 (low noise amplifier), 2 mixers 730, 2 analog summers 740 and 2 A / D 750 (analog to digital converters, which may advantageously use sigma-delta ADCs). Not shown are the same LNAs and mixers for the other 3 receive elements in the cell labeled RX1, RX2 and RX3. The analog receive function down-converts the received signal to baseband for the following reasons:

[0052] The four channels can be combined after very simple phase adjustment of the local oscillators, so each cell requires only 2 ADCs instead of 4. Compared to the RF approach, the me and Q ADC bandwidths are reduced by a factor of f0 / Δf, where Δf is the bandwidth of the transducer and f0 is its center frequency.

[0053] The digital beamforming logic only needs to serve 128 data channels instead of 512 data channels.

[0054] In one implementation, the analog transmission section uses 128 input signals with correct timing from the digital block. The analog transmission section consists of a level shifter 760 (for shifting the output voltage to the appropriate voltage for the transmission element) and an output stage 770. In an embodiment, the output stage 770 supports a three-level output of +V, +V / 2 and GND to supply a symmetrical waveform around +V / 2. This ensures that the transmission circuit actively returns to a resting voltage of +V / 2. In an implementation, the "ground" electrode shared by all MEMS elements is maintained at ground or a bias voltage, which is selected to ensure that the field across the piezoelectric film will never reverse and is always the same as the polarization direction of the piezoelectric film. This optimizes the reliability of the film. In one implementation, the operating frequency varies between 0.5MHz and 3.0MHz. Some examples of the signals sent include:

[0055] • Channel mode operation: short and symmetrical 1-cycle or 2-cycle tone burst at maximum output voltage to be used for skull characterization.

[0056] Beam mode operation:

[0057] o For grey scale, short pulses centered at a lower frequency in case of receiving the second harmonic via eg pulse inversion signaling.

[0058] o For cranial vascular mapping, a tone burst of 4-8 cycles centered at the mid-band frequency.

[0059] Stimulation Mode Operation: Long tone bursts (100-5000 cycles) centered between 0.5 MHz and 2.0 MHz.

[0060] Typically, the rise and fall times of the transmit waveforms are matched, and the delay amount for transmit beamforming is approximately 75ns.

[0061] Digital Blocks

[0062] The functions of the digital block are:

[0063] Provide a "dynamic" delay to the received signal, i.e., a delay that varies with the depth of the sound return. This delay needs to take into account how the element group is combined in the simulation. Since a fractional sample delay is required, delays are applied in the coarse and fine stages. In one implementation, CORDIC (Coordinate Rotation Digital Computer) rotations are used for the fine delay. In a different implementation, a decimation filter is used in which the start and stop addresses are adjusted.

[0064] • Control of the signal amplitude as a function of depth for each channel.

[0065] Sum the channel data after the above delay and amplitude processing.

[0066] Provides adjustments for delay and amplitude to compensate for the aberrations of the skull.

[0067] The ability to accept digital data from other ASICs, delay it, and sum it with the beamforming data generated from the receive elements. This enables the slices to be "chained".

[0068] Create digital signals with correct timing for each analog transmitter. Delay these digital signals relative to the start signal by an amount that has a geometric component based on skull measurement and a correction component.

[0069] • The ability to operate in the following modes: (1) normal beamforming mode, where channels are combined after delay and amplitude adjustments; (2) channel mode, where data is not beamformed; this “channel” mode helps characterize the skull; and (3) stimulation mode.

[0070] Figure 8 An exemplary functional block diagram of the digital block 620 is shown. At the heart of the digital block 620 is the delay weight element (DWE) 810 subsystem. Fig. 9 This subsystem, which is further detailed in , creates the correctly timed signals to be fed into the transmitter analog functions. The receive section receives the data fed to it from the baseband ADC output and performs digital operations, resulting in beamforming of the data. Figure 8 , in an embodiment, the digital block 620 includes 128 delay weight elements 810, an interconnect 820, a channel adder tree 830, a FIFO and beam adder 870, 16 LVDS transmit (LVDS TX) 860 and 16 LVDS receive (LVDS RX) 880 pairs, a QSPI slave interface 850, and a global logic block 840. The global logic block is responsible for (among other functions) taking the representation of the beam direction and focus parameters supplied to it from the beam list and calculating the low-level parameters stored in memory 960 and memory 940.

[0071] LVDS transmit 860 and LVDS receive 880 are used for slice chaining. The received LVDS data is delayed in FIFO 870, then added with the slice's data in beam adder 870 and sent to the next slice in the chain via LVDS TX 860.

[0072] Analog data is received from analog block 610 and a dynamic delay as a function of depth is applied by delay weight element 810. The delay weight element also adjusts the amplitude of the channel data as a function of depth again. After the amplitude and delay of the channel data are appropriately adjusted, it is fed to the channel adder tree 830 for summing. The QSPI slave interface 850 provides an interface to the RT control device 210. The RT control device 210 uses this interface to provide beam lists and other control information. The global logic block 840 provides overall control to the digital block 620, and the interconnect 820 provides on-chip interconnection between the various functional blocks.

[0073] Delay Weight Element (DWE)

[0074] Fig. 9 An exemplary block diagram of a DWE 810 is shown. The DWE 810 includes a phase calculator 950. The transmit delays specified in the beam list are implemented in a TX delay circuit block 955. These delays may be implemented using memory or by other techniques.

[0075] Dynamic receive processing is more complex because the delays and weights must vary with depth. Values ​​of delays and amplitudes to be applied to data at certain depths are stored in memory (RX delay storage) 960 and memory (RX weight storage) 940. The logic in blocks 945 and 935 allows interpolated delays and weights to be applied at depths between the depths specified in memory 960 and 940.

[0076] The signal path of the delay weight element begins at the digital filter and extractor 910. Here, the ADC output is filtered to remove out-of-band noise, and the sampling rate is reduced to the minimum value consistent with sufficient sampling of the baseband data. Thereafter, a coarse delay is achieved by the write address and read address of the RX delay memory 915. The generation of the read address that varies with depth is completed by the read address generation block 920. Fractional delays can be achieved using the RX delay interpolation logic 945. Further fine control of the signal phase can be achieved using the phase adjustment logic multiplier 925. After this process, the signal received from the point target will be time-aligned across all channels and the signal will be ready for summation. Before summation, the weight multiplier 930 implements apodization or weighting operations. This is driven by the RX weight interpolation logic 935 according to the data in the memory 940.

[0077] Closed-loop TUS method

[0078] Fig.10 1 is an exemplary method 1000 for closed-loop operation of a TUS wearable device 120. The method 1000 may be performed using the system 100. The method 1000 may correct for skull aberrations, determine whether the device 120 is placed in the correct position, and whether the device has been displaced or moved during use. Fig.10 , in operation 1010 , the wearable device 120 is placed on the scalp of the subject 110 .

[0079] In operation 1015, various operating parameters such as a beam list are loaded into the wearable device. For example, the RT control device 210 loads the beam list from the memory 220 and provides it to the ASIC 260 together with the control information.

[0080] In operation 1020, the device 120 is operated in a "channel" mode for skull characterization. As previously described, various characteristics of the human skull complicate the delivery of ultrasound to the desired target. In this operation, the method characterizes the effect of the skull on the ultrasound of each transducer element. In this operation, the transducer elements of the TUS device transmit short pulses with a suitable center frequency. The received data is captured with RX beamforming turned off, and the data is used to characterize the skull of the subject.

[0081] In operation 1025, the data received and sent from operation 1020 is analyzed to determine the effects of the skull and appropriate correction parameters are calculated for each transducer element. The beamforming parameters for all transducer elements are updated.

[0082] In operation 1035, the method maps and aligns the brain structure (or anatomical representation) with the device operating in "beam" mode. First, a volumetric grayscale image (2-D image) of the brain structure is acquired, and then volumetric vascular data (blood flow) is acquired using Doppler processing.

[0083] In operation 1040, the information acquired in operation 1035 (brain structure and volumetric vascular data) is determined or labeled using pattern recognition, machine learning, or any suitable algorithmic approach. This helps determine whether the device is properly placed on the subject's scalp and allows stimulation to be targeted assuming adequate positioning.

[0084] In operation 1045, the method determines whether the stimulation target is reachable based on the location of the device 120 on the scalp of the subject 110. If the target(s) are not reachable, the next operation in the method is operation 1050. If the target(s) are reachable, the next operation is operation 1055.

[0085] In operation 1050 , a message is sent from the device 120 to the subject 110 to relocate the device. For example, the message via the secure network device 130 may be displayed on an application on the subject device 140 . The method proceeds to operation 1015 .

[0086] In operation 1055, stimulation parameters are loaded. In the healthy mode, the subject is provided with various preset stimulation waveforms. In the medical treatment mode, the clinician will specify the stimulation waveform and other parameters such as treatment length, burst length, time between bursts, etc. For example, the real-time control loads the beam list from the memory 220 and provides it to the ASIC 260 along with the control information. The steering and focusing information for all TX transducer elements is calculated. This includes the delay set for each transducer element and potentially includes parameters for implementing advanced beamforming involving plane wave transmission at different angles.

[0087] In operation 1060, stimulation is started. The stimulation waveform is a long tone burst (300 to 5000 cycles) centered at a suitable stimulation frequency interspersed with periods of inactivity. Stimulation waveform parameters such as tone burst duration, pulse repetition frequency, acoustic frequency, amplitude intensity, etc. are determined by the clinician's treatment plan (or prescription). The device 120 ensures that no stimulation waveform parameters exceed a safe range and that the subject 110 cannot change the stimulation waveform parameters to an unsafe level. The device 120 periodically acquires grayscale and vascular data. For example, the device can acquire grayscale and vascular data after every 1 / (pulse repetition frequency) or a multiple of 1 / (pulse repetition frequency). Periodically acquiring grayscale and vascular data helps determine whether the headband 120 is displaced during use.

[0088] In operation 1065, a cross-correlation algorithm or other algorithm is used on the sequentially acquired grayscale and vascular data to determine whether significant relative motion between the device or headband 120 and the scalp of the subject 110 occurs. Significant relative motion between the device and the scalp of the subject may result in ineffective stimulation because the stimulation beam may no longer be directed at the target tissue. If motion is sensed, the next operation is operation 1070. If there is no significant movement, the next operation is operation 1075.

[0089] In operation 1070 , the steering and focus information for all TX transducer elements is recalculated. Method 1000 proceeds to operation 1060 .

[0090] In operation 1075, method 1000 checks whether stimulation is complete. Stimulation may end due to an explicit command (from the subject), or stimulation may end because a specified NTB (number of tone bursts) has ended. If stimulation is not complete, the method proceeds to operation 1060; otherwise, the method ends.

[0091] CROSS-REFERENCE TO RELATED APPLICATIONS

[0092] This application claims the benefit of priority to U.S. Provisional Patent Application Serial No. 63 / 601,577, entitled “WEARABLE CLOSED LOOP TFUS SYSTEM,” filed on November 21, 2023. The subject matter of this related application is incorporated herein by reference.

Claims

1. A transcranial ultrasound system, i.e., a TUS system, comprising: A headband device comprising one or more transcranial ultrasound (TUS) pads, wherein each of the TUS pads further comprises a plurality of transducer elements connected to an integrated circuit; and A computing device is configured to wirelessly communicate with the headband device, wherein the computing device generates a stimulation signal specification and sends the stimulation signal specification to the headband device, wherein the TUS patch outputs the stimulation signal via the headband device.

2. The system according to claim 1, wherein: The TUS sheet also includes a matrix of TUS sheets.

3. The system according to claim 1, wherein: The headband assembly also includes at least one physiological measurement subsystem.

4. The system according to claim 1, wherein: The computing device performs authentication, encryption, tamper detection, and side-channel attack protection on behalf of the headband device.

5. The system of claim 1 , further comprising a real-time control circuit coupled to the headband device, wherein: The real-time control circuit reads the beam parameters from the memory, so that the TUS chip outputs the beam according to the beam parameters.

6. The system according to claim 5, wherein: The real-time control circuitry causes the headband assembly to operate in a channel mode, a beam mode, or a stimulation mode.

7. The system according to claim 6, wherein: In the channel mode, the real-time control circuit causes the headband device to obtain skull characterization data using the TUS patch, wherein the computing device generates skull compensation parameters based on the skull characterization data.

8. The system according to claim 7, wherein: In the beam mode, the real-time control circuitry determines whether the headgear assembly is aligned with a desired anatomical structure based on the skull characterization data and the anatomical characterization data.

9. The system according to claim 7, wherein: In the beam mode, the computing device detects blood flow or blood vessel location based on the skull characterization data.

10. The system according to claim 7, wherein: In the channel mode, the real-time control circuit causes the TUS tile to communicate with an adjacent tile in a tile array, and wherein the real-time control circuit disables transmit beamforming or receive beamforming.

11. The system according to claim 6, wherein: The real-time control circuit enables the headband device to output a stimulation beam.

12. The system of claim 1, wherein: The TUS sheets are arranged in an array and are electrically coupled to adjacent sheets in the array.

13. The system according to claim 1, further comprising a memory for storing control parameters for the headband device, wherein: The control parameters define beam parameters for a plurality of operating modes.

14. The system of claim 1, wherein: The TUS sheet includes a micro-electromechanical system device, namely a MEMS device, and the MEMS device includes at least one of a piezoelectric micromachined ultrasonic transducer, namely a pMUT, and a capacitive micromachined ultrasonic transducer, namely a cMUT.

15. The system of claim 14, wherein: The MEMS device further includes at least one receiving portion and a transmitting portion, wherein the transmitting portion is electrically separated from the at least one receiving portion.

16. The system of claim 15, wherein: The MEMS device is electrically coupled to an integrated circuit.

17. The system of claim 16, wherein: The integrated circuit includes an analog block and a digital block.

18. The system of claim 17, wherein: The analog block receives data from at least one receiving portion of the MEMS device and transmits data to the transmitting portion of the MEMS device.

19. The system of claim 18, wherein: The integrated circuit includes at least one of an LNA, a summer, a Σ-Δ ADC, a level shifter, and an output stage.

20. A method for a transcranial ultrasound system (TUS system), comprising: placing a headband device on the scalp of the subject, wherein the headband device includes a plurality of transcranial ultrasound sheets (TUS sheets); Obtaining TUS parameters from a memory, wherein the TUS parameters are used to specify an operation mode of the TUS slice; obtaining skull characterization data from the TUS slice based on the TUS parameters; calculating skull aberration correction parameters based on the skull characterization data; adjusting the TUS parameters based on the skull aberration correction parameters; as well as A stimulation signal is generated based on the adjusted TUS parameters, wherein the TUS patch transmits the stimulation signal via the headband device.

21. The method of claim 20, further comprising generating volumetric grayscale data or vascular data to visualize skull bones or blood vessels.

22. The method of claim 21, further comprising verifying positioning of the headband assembly based on the volumetric grayscale data or vascular data.

23. The method of claim 22, further comprising transmitting the stimulation signal in response to verifying positioning of the headband assembly.

24. The method of claim 21, further comprising: determining, based on the volume grayscale data or the blood vessel data, whether a relative movement exceeding a threshold value has occurred between the headband device and the subject's scalp; as well as In response to the determination, a message is generated to reposition the headband assembly.

Images